Today, it is little known that for forty years at the height of the British Empire, dummy pawns were the scourge of tournament play, and even grandmasters ran scared. (Possible exaggeration here.)
The heresy raged from 1862-1904. See Eminent Victorian Chess Players: Ten Biographies by Tim Harding. However, the origins of the dummy pawn goes back to 1847 or before.
Timeline of known events
(+/- indicate events pro/con dummy pawns)
1847 + the earliest known mention of the dummy pawn occurs in the American Chess Magazine
1851 + arrival in London, first appearing in the “The Chess Player’s Chronicle,” and called “The Revolutionary”
1854 - an example by Alexander Petrov appears in “Règles du jeu des échecs” in which it was denounced
1860 + Howard Staunton in his "Chess Praxis" wrote "no one will consider [dummy pawns] worthy of serious consideration."
1862 + International London Congress - provides rule for dummy pawns
1865 - William Pavitt, UK, explores the effects on chess composition in “The Chess World”. According to him: "an exquisite stroke of chess legislation," but condemns it.
1870 + Baden Congress - takes same line as London Congress of 1862.
1876 - new edition of Staunton's "Chess Praxis" - his rules still say no dummy pawns
1879 - American Chess Journal - "a frivolous and undesirable innovation"
1880 + Fred Thompson, in “Pawn on 8 In Chess Problems,” advocates for them as aids in chess composition
1883 - London International Chess Congress banned dummy pawns
1883 + Francis C. Collins composes a s#2 allumwandlung for “The New Zealand Mail”
1889 + Steinitz - supports London 1862 decision, and reports that so have many Congresses
1894 + "The Chess Pocket Manual" - "famous Dummy Pawn rule"
1898 + a spectacular Sam Loyd #3 composition
1900 + a superb #4 by E. J. de Savornin Lohman
1895 - the legendary Hastings congress banned dummy Pawns banned
1903 - unsupported statement that the reign of dummy Pawns ended in this year ([http://www.chess-poster.com/english/history/the_pawn.htm])
1904 - founding of British Chess Federation, which according to Harding published the rules that finally exiled the monstrous dummy pawns to history
1913 + leading problemist Thomas R. Dawson composes a classic study with 3 dummy pawn promotions by White
1996 + John Beasley publishes a dummy pawn promotion study in EBUR
2017 + dummy pawns gains new life in chess.com and Matplus.net forums
2018 + chess.stackexchange.com diagram editor (still) supports dummy pawns
2020 + Popeye & Jacobi problem engines (still) support dummy pawns
See dummy pawns in action in our own heretical diagram editor below! (Except for the Ponziani game which is even more bonkers, and cannot therefore be shown.)
1847: Misty Beginnings In America
The origin of dummy pawns is mysterious and unknown. It most likely occurred in the early 19th century though.
The earliest known mention of the dummy pawn is from “The American Chess Magazine” in the year 1847. On page 113 the following problem was shown with the attached question.
[Title "F. Leake - American Chess Magazine 1847, Can White force checkmate in five moves? "]
[FEN "2B5/8/1K1k4/4pP1N/3pP3/Pp1P4/p7/7R w - - 0 1"]
It says: “ A question of great importance - one which is not provided for in the laws of the game of Chess, and to decide which we know of no precedent- is here involved. Can White force Checkmate in five moves? That's the question."
If Black is obligated to promote to a piece, then there is a mate in 5: 1. f6! b2 2. Rb1 a1=B 3. Rxb2 Bxb2 4. Ng4/g7 ~ 5. Nf5/e8#. However, if Black may keep it as a pawn, then there is no mate in 5. There is a unique mate in 6 though: 1. f6! b2 2. Rb1 a1=Pawn 3. Kb7 Kc5 4. f7 Kb5 5. f8=Q Ka5,Ka4 6. Qb4#
On pages 136-139 appears correspondence giving opinions as to whether or not there is a mate in 5.
No: “(Skipped text) I presume that the question you raise is this: When a Pawn reaches the eighth square, is the player obliged to change it at once for a Piece, or has he the option of allowing it to remain dead on the eighth square? In my opinion, he is obliged to change it at once for a Piece ; he has no option but in the choice of the Piece so exchanged. The Twenty-third Law of Chess, according to Walker's Treatise, (1842,) is :— Directly the Pawn attains the eighth square, or extreme rank of the board, it must be replaced by a Queen, Rook, Knight, or Bishop, at the option of its owner; and this without regard to whatsoever Pieces he may already have on the board. Should you not replace the Pawn with a Piece, before your opponent move, he may take it off the board as forfeited. (Skipped text). Though you appear to think this a very nice question, I have but little doubt as to the nature of your decision, should you think proper to make one. And with all due deference to your well-earned reputation as a Chess-lawyer, I will venture to risk the opinion, that the " big wigs" in London will be surprised that you should regard the question as one unprovided for in the laws of Chess." - Philo-Chess
Yes: “(Skipped text) For myself, however, I incline to a different opinion. The Checkmate in this case cannot be forced in five moves, unless the Pawn, on reaching the "royal line," loses its character as a Pawn, and immediately and necessarily becomes a Piece of some kind; or unless Black is compelled, by some law of the game, to replace the Pawn at once by some Piece. (Skipped text) It is obvious that the right of promotion is granted as a favor or privilege, and that upon the assumption that a Piece is of greater value than a Pawn. Why, then, limit this privilege, by confining the choice to a Piece? Why, especially, require the player, who is entitled to a reward for his skill and success in conducting his Pawn to the "royal line," to submit to a positive disadvantage, by compelling him to exchange it for a Piece of inferior value to him? Finally and conclusively, the rule, strictly construed, is anti-republican-and this is a "free country!" Why should the Pawn have "greatness thrust upon him If he is satisfied with untitled merit, why not permit him to indulge his tastes for private station? Why elevate him, against his inclination, to a rank to which he does not aspire?“ - A.D.P.
The wonder is that this problem is much more sophisticated than any other dummy pawn problem that would be made for nearly two decades, until Pavitt came along with his 5-mover in 1865.
1851: A Start In London
The dummy pawn, 4 years later, made an appearance on the other side of the Atlantic Ocean in what is a remarkable coincidence. Edward Winter managed to find the first mention of a dummy pawn in England in C.N. 7619. In the September 20, 1851 edition of “The Chess Players Chronicles,” the below problem called “The Revolutionary” appeared.
[Title "Josef Kling & Bernhard Horwitz - The Chess Player 1851, Revolutionary Question XIX: How can White draw?"]
[FEN "r7/1Pp5/2P3p1/8/6pb/4p1kB/4P1p1/6K1 w - - 0 1"]
The solution, which was given in a week later, stated: “The solution of No. 19 is Pawn takes Rook, remaining a pawn. It may seem ridiculous to many chess-players, but we think candidly that, when first the rules were laid down, nobody thought of such a case, or else the rules would be a little more precise on that point."
In the correspondence section of the 9/20 issue the editors also said “ We are of opinion that a pawn, when reaching his eighth’s square, is entitled to become any piece the player thinks best; and, naturally enough, if it prove disadvantageous to become a piece, it may still remain a pawn. (See Diagram 19.) This question among others, we thought, should have been discussed at the Tournament.”
1854: Spread In Europe & Rejection
From 1851 onward, the idea of the dummy pawn seems to spread across Europe. One indication is the appearance of another dummy pawn problem, this one also being White to play and draw, that was created by the Russian composer Alexander Petroff and published in the 1854 French book Règles du jeu des échecs. On page 97, the problem was given.
[Title "Alexander Petroff - Règles du jeu des échecs 1854, White to play and draw"]
[FEN "rn6/pp2n3/2p5/P4k2/8/6p1/6Pp/7K w - - 0 1"]
1. a6 Nd5 2. axb7 Nf6 3. bxa8 null
However, this problem is flawed because Black can play 2... Nf4!, 3. bxa8 Nxg2 4. Kxg2 Kg4, and Black wins. White doesn't win even if they promote-3. bxa8=Q Nd3 4. Qxa7 Nf2+ 5. Qxf2 gxf2 6. g4+ Kxg4 winning.
The French text, roughly translates, reads: “This subject, we must say that Mr. Petroff communicated to us the following problem, in which, supposing the best possible blows on the part of the blacks, the whites could not postpone the game, unless they were allowed to refuse any promotion to their pawn entered lady, that is to say to leave it pawn. (Diagram of the problem.) On this occasion, Mr. Petroff asks us to submit to the judgment of the European authorities in matters of gambling, if it would not be advisable to introduce into the regulations a clause which would allow the possessor of a pawn entered to queen, to renounce to any increase in its forces, leaving this pawn inactive and immobile, as a pawn, until it is taken or until the end of the game?
“For its part, the Society of Saint Petersburg Chess Lovers did not believe that it could accept this opinion (expressed in the form of doubt by Mr. Petroff himself), contrary to the fundamental principle of the game, which 'does not admit the existence of pawns to the strips of the chessboard originally occupied by the pieces.'"
This shows that the dummy spread like wild fire, relatively speaking for the times, and within 3 years it was already an outrage in at least France and Russia. It would take awhile for the English to get their turn.
1860: Staunton, bastion of orthodoxy
The "dummy pawn rule" was not part of the official rules up to 1860. See "Chess Praxis" by Howard Staunton 1860. Staunton was a very influential player and author, and in this book codified the Laws of Chess, in a way which is still recognizable in the style of the FIDE Laws today. On page xi, in the Introduction he lists chronologically the many antecedents of his work from Ruy Lopez in 1561, to the 1850s.
*Page 6, Queening a Pawn: “When a Pawn has reached the eighth or last square on its file, it immediately assumes the name and power of any Piece its player may select, except a King, whether such Piece have previously been lost or not; and, if the player does not select a Piece, such Pawn is always to be considered a Queen.”
On page xii, he had defined "Piece" to explicitly exclude pawn. The real fun comes later:
Chapter 4, "A Queened Pawn: “On this subject great difference of opinion prevails, but the rule laid down in the text is that universally observed in practice. In the middle ages, as we have previously explained, the Queen could only move one square at a time diagonally. She was, therefore, by far the weakest piece on the Board. And the ancient law, which required that every Pawn pushed on to the eighth square should become a Queen, was really a restrictive enactment, since it gave the promoted Pawn as little additional power as the conversion could offer. But the spirit of the modern game is to regard the Queening of a Pawn as the highest feat a player can accomplish, and to reward it with the greatest possible advantage. So that a player in the present day is not only allowed to select a second or third Queen with its enormously extended power, but may choose any other Piece that would be more advantageous in a particular position. Various modifications of this law have existed in different places and at different times. It has been held, for example, that the Pawn should only acquire the power of the Piece on to whose square it had been played, or of a Piece already lost. It has also been proposed to limit the conversion to either a Knight or a Queen, as the latter comprises the power of Rook and Bishop; and the Pawn has sometimes been required to perform certain additional moves before becoming entitled to the privileges of a Piece. The most plausible of the regulations is that the Pawn should only supply the place of a Piece already lost, so as to avoid, what Philidor so violently denounced, plurality of pieces. But then comes the difficulty of providing for the case where a Pawn has reached the eighth square before any Piece has been lost. It has been proposed to leave such a Pawn, as it were, dormant, until a capture should have made among the Pieces a vacancy for it to supply. Ponziani, an advocate of this regulation, has given the following little game as an example of it.”
Example of non-standard castling & promotion
[Title "Ponziani scheme - White to move"]
[FEN "Bn1qk1nr/p1p2ppp/8/8/4P2b/5N1b/PPPP3P/RNBQ1rK1 w k - 0 1"]
1. e4 e5 2. f4 exf4 3. Nf3 Be7 4. Bc4 Bh4+ 5. g3 fxg3 6. Kh1&Rf1 d5 Italian style castling!
7. Bxd5 Bh3 8. Bxb7 g2+ 9. Kg1 gxf1 without promoting
10. Bxa8 Bf2# & f1=R# only now does pawn f1 promote to a rook
“In this game the black Pawn, which attained the eighth square on its ninth move, neither gave mate nor check, because the player had lost no Piece into which it could be converted, and it was not in a situation to check as a Pawn. But as White chose to capture the Rook on his tenth move, Black plays down the Bishop with one check, giving another and mate, with the Pawn now become a Rook in place of that captured. There is so much of the absurd about such a finale as this that no one will consider it worthy of serious examination, and it is only mentioned as a matter of curiosity.”
1862: The Madness Begins
As described in the Edward Winter link in another answer, the dummy pawn rule was codified in the International Congress of London in 1862, despite Staunton's outrage. For a generation, the war raged between the two camps, echoing the chaos in the European political space at this time. English chess papers would bicker about it for decades to come, with numerous joke problems and articles, with most condemning the dummy pawn.
1865: The Chess Composers Also Grumble
In 1865, 3 years after the passing of the law allowing the dummy pawn, an early article by William Pavitt against the dummy pawn appeared in the “The Chess World”.
Pavitt gave this simple mate in three he created that worked under normal promotion rules
[Title "William Pavitt - The Chess World 1865, Mate in 3"]
[FEN "K7/2N5/2p5/k1P3R1/p4N2/P2P4/6p1/8 w - - 0 1"]
1. Ng6! g1=Q 2. Ne5 Qc1 3. Nxc6#
However, in essence, with the dummy promotion rule, the problem wouldn’t work as there is no mate in two. He showed a mate in 5 to show how the problem would have to be corrected to conform with the dummy pawn. (Note that this is a correction the original created by James Malcom as the original is incorrect).
[Title "William Pavitt (corrected J. Malcom 2020), Mate in 5"]
[FEN "8/K1N5/2p5/k1P5/p7/P2PP3/3p2pN/3B4 w - - 0 1"]
1. Ng4 g1 2. Bf3 d1 3. Bd5 cxd5 4. Ne5 d4 5. Nc4#
Pavitt showed these problems to give examples on how allowing the dummy pawn would greatly affect chess problems, and he condemned them as a ruiner of chess problems.
Most chess composers in the chess problem world were eventually set against the dummy pawn due to how it would interfere with known chess problem and the needless work that would be needed to incorporate it in. An occasional exception, i.e. Christmas and trick problems that intentionally used it, was generally accepted for the humor. In fact, many problem tournaments specified to rules to disallow the dummy pawn.
A typical data point from this dark time is the Baden Chess Congress of 1870, which adopted the 1862 rules (source). An article covering this begins attractively:
“Unmoved by the shrill notes of the war trumpets sounding in their ears, the Chess players have commenced their mimic strife in Baden.“
In the mean time, Staunton came out with another edition of his Praxis (source). He did not mention the 1862 rules, but he did give the previously discussed 1854 example by Petroff.
1880: A Rare Friend To The Dummy
However, not everyone gave disdain to the dummy problem. A few chess composers thought that it should be allowed. An old article by Fredrick Thompson, published in 1880 in “Chess Player's Annual And Club Directory,” called on “Pawn On 8 In Chess Problems” gives an insight into the supporters of the dummy pawn. In his article. Thompson advocated for the dummy pawn to be allowed in chess problems.
Various reasons are given, the most interesting perhaps being this: “ The main objection to P on 8 is, that it breaks a fundamental law of Chess. So it would if pawn stood on square 1, but on square 8, in the enemy's territory, is quite another thing. As Staunton in his Praxis says—" The ancient idea supposed the Chess Board divided into two belligerent kingdoms." Have common soldiers never mounted guard in the palace of a foreign potentate at the conqueror's behest? Ought pawn to be played to 8 at all if he may not remain there? His captain bids him stand sentinel as a private soldier, and it is not for the discomfited foe to say him nay. Ought pawn to attack a piece or check the King from square 7, if he has no business on 8?”
The logic is that if the pawn has power on the 7th rank as a pawn, why should can it not extend its rights as a pawn on the 8th rank?
A more straightforward reason is also given: “The present pawn-queening law bristles with innovations upon ancient usage. As Mr. Staunton wrote on March 18, 1865—"The old laws from which you quote, have been constantly modified of late years." Fundamental law, forsooth! The law never had a bottom. No one will deny the use of pawn on 8 in problems if pawn on 8 is permissible in play. The " Competent authority " for the laws of Chess must be players, not problem composers, for the laws, of the game govern the problemist, not the problemist the laws of the game.”
Lastly, he advocates for it as to prevent cooks in chess problem, to make possible otherwise impossible ideas realizable at last. He gives a problem to illustrate his defense of the dummy pawn, since it would be cooked without the pawn on c8, since without there would be a cook starting with 1. Qc8+.
[Title "Fredrick Thompson, Mate in 3"]
[FEN "2PQ4/6p1/3P2B1/3Pk3/8/2N1N1P1/8/7K w - - 0 1"]
He ends with this: “ I claim, therefore, P on 8 in Chess strategy, and conclude this paper with a problem commenced in July, 1879. (Skipped text). Dropping the men one square down the board also produces a cook, in fact the dead pawn saves the problem from being relegated to the waste basket. I consider this one of the best compositions I have put together during a practice of thirty years, and I fail to see why it should be treated to Lynch-law, when the queening-a-pawn rule is to "reward it with the greatest possible advantage."
1883: London Congress
Another data point is the London International Congress of 1883, where the rules include:
“10. A Pawn reaching the eighth square must be named as a Queen or Piece, at the option of player, independent of the number of Pieces on the board. The created Queen or Piece acts immediately in its new capacity. Until the Pawn has been so named, the move is incomplete.”
In almost all other cases in these rules, Pieces and Pawns are taken to be distinct, but this is the first occurrence I've ever seen a usage of Piece to exclude the Queen! However, I think it is clear that dummy pawns are on the retreat.
1883: Far Reaches-The Dummy Pawn’s Travels
One example that can be found of how far the dummy pawn reached is a problem from the 3/10/1883 edition of the “New Zealand Mail.” New Zealand was then part of the British Empire, is so not so surprising that the dummy pawn got that far, but it is pretty far though!
[Title "Francis C. Collins - New Zealand Mail 3/10/1883, Selfmate in 2"]
[FEN "1n3R1n/3p1p2/B1pP1Pp1/2B1K1P1/2r5/3k4/NR2p1Q1/8 w - - 0 1"]
The point is that capturing the b8 knight fails because Black can then promote to a pawn and there is no selfmate after that. The solution is 1. Rxf7!, such that after e1=P, the move. 2. Rxb8 forces 2... Nxf7#.
The solution, printed in the 4/21/1883 edition, says this: "This position has been composed by Mr. Collins in accordance with the British Chess Association’s 13th Law: —“When a pawn has reached the eighth square, the player has the option of selecting a piece, whether such piece has been previously lost or not, whose name and powers it shall then assume, or of deciding that it shall remain a Pawn.” This rule of the “dummy pawn” has never met with much favour amongst chess-players, and there are very few clubs or societies which observe it. We have received no correct solution to the problem (which is a most ingenious one), the result doubtless of the non-acquaintance by our friends of such a rule as the foregoing in any of the codes of chess laws."
This shows just how weak of an entity the dummy pawn has become and how very few looked upon it with favor.
1895: Hastings Congress
By the time of the famous Hastings Congress in 1895, dummy pawns are clearly gone from serious play. The Congress rules adopt exactly the same wording as London 1883. (All attendees were searched to make sure no dummy pawns were being smuggled into the venue.)
1898: Sam Loyd #3
In what is widely known as the last hurrah, no discussion of the dummy pawn period would be complete without the following beautiful problem by the American composer Sam Loyd.
[Title "Sam Loyd - American Chess Magazine, Mate in 3"]
[FEN "N1Br4/2Pb1P2/3k4/1P2R3/1P2K3/B7/8/8 w - - 0 1"]
(1. cxd8=Q/R? stalemate)
(1. cxd8=B? Bf5+)
(1. cxd8=N? Bc6+)
threatening 2. f8=Q/B#
(1... Bf5+ 2. Rxf5 Ke7 3. f8=Q#)
(1... Bc6+ 2. bxc6 Kxc6 3. b5#)
(1... Bxc8 2. f8=Q+ Kd7 3. Qe7#)
(1... Bxb5? 2. Re6#)
1900: An Ingenious #4
But in fact Loyd's was only the penultimate hurrah. Two years later the brilliant but little-known E. J. Lohman produced this gem, which is as ingenious as Loyd’s #3. This is what should be known as the last hurrah of the dummy pawn, the pinnacle of the times when you could make a position and legally solve it by promotion to a dummy pawn.
[Title "E. J. de Savornin Lohman - De Amsterdammer 02/09/1900, Mate in 4"]
[FEN "2B5/1pR1P3/1P1k4/8/2K5/5P2/8/8 w - - 0 1"]
1. e8! Ke5 2. Rf7 Kd6 3. f4 Kc6 4. Rf6#
(1. e8=Q/R? stalemate!)
(1. e8=N+? Ke5 2. Rf7 stalemate!)
(1. e8=B? Ke5 2. Rf7 Kd6 3. f4 stalemate!)
1904: The Mischief Ends
But it was only with the arrival of the British Chess Federation that the final nail in the coffin was driven home. There is still some lack of clarity as to the precise year this happened, but it was around 1904.
1913: Last Call
Nearly a decade after the dummy pawn was finally killed off in law, the famous fairy chess composer Thomas R. Dawson took some time to make a drawing study with it. White must create three dummy pawns in order the achieve the draw via self-stalemate!
[Title "Thomas R. Dawson - The Chess Amateur 1913, White to move and draw"]
[FEN "6K1/6PP/3p2PP/3P1ppp/8/ppp5/qrp5/1rk5 w - - 0 1"]
1. Kh8 Kd2 2. g8 Re1 3. g7 Re6 4. dxe6 Rb1 5. e7 b2 6. e8 Qxg8+ 7. hxg8 null 8. h7 null
Afterward, the dummy pawn slowly faded into history, only mentioned now and then, until the dawn of the Internet.
1996: A Lone Bright Spot
In 1996, the dummy pawn had another brief period of relevance when John Beasley made a small study in which the only way for White to win is to promote to a pawn. It is the first known dummy problem in which White wins since the Loyd #3-nearly a century! I will leave to you figure out the reasons to avoid promoting.
[Title "John Beasley - EBUR 1996, White to move and win"]
[FEN "5r1K/2k1P3/3N1Bqp/N2Q1B1P/8/8/8/1R6 w - - 0 1"]
2017: Just For Fun
The Internet has allowed many to create and share their own dummy pawn problems. While there is no doubt many problems been created, perhaps the pinnacle of achievement was when Peter Wong created an entirely new problem in which White wins by promotion to a pawn!
The problem was shared on the Matplus.net forums in November, 2017, although. the composer has originally shared somewhere on chess.com.
The chess composer Geir Sune Tallaksen Østmoe managed to add a golden touch to Peter Wong in that it is in both sides best interest to promote to a pawn!
[Title "Geir Sune Tallaksen Østmoe & Peter Wong - Matplus 11/12/17, Black to move and White to win"]
[FEN "5qnk/3pPB2/1p1p1N2/1pbK1NP1/1p1pP3/1P1P4/6p1/7R b - - 0 1"]
1. gxh1! exf8!
White easily wins with an extra piece. The full solution and the refutations of each promotion can be viewed here.
Also from the same forum, here are two possible 6-piece settings for dummy pawn draw:
[Title "ichai - MatPlus 7/2/2020, White to move and draw"]
[FEN "8/2P5/8/8/8/2r5/p1r5/K1k5 w - - 0 1"]
(1. c8=Q? Rb3)
And the current economy record-holder:
[Title "A.Buchanan - MatPlus 13/5/2020, White to move and draw"]
[FEN "7K/5rPp/7k/8/8/8/6p1/8 w - - 0 1"]
(1. g8=Q? g1=Q! 2. Qxg1 Rf8+ 3. Qg8 Rxg8+ 4. Kxg8 Kg6!
(2. Qxf7 Qd4+ 3. Kg8 Qd8+ 4. Qf8+ Qxf8+)
(2. Qf8+ Rxf8#)
(2. Qg7+ Qxg7#)
(2. Qxh7+ Rxh7#)
(2. Qg5+ Kxg5)
(2. Qg6+ Kxg6,hxg6))
Will the dummy pawn rise again?
It seems unlikely that FIDE will introduce the dummy pawn for over the board play, although it would be cool if they did. For compositions, dummy pawn promotion, as the ultimate under-promotion, is paradoxical and dramatic: exactly what a problem chess move should be. However, introducing a 5th event at promotion would mean that the vast range of allumwandlung studies, including solutions to the famous Babson task, would suddenly only be 80% complete. This seems unlikely to happen.
Where Can I Find Dummy Pawn Problems?
Currently, the largest known catalogue of dummy pawn problems is on the Die Schwalbe Chess Problem Database Server, known as the "Schwalbe PDB for short. Searching "K='dummy pawn' in the search function currently brings up a list of 60 dummy pawn problems, from across the 19th-21st centuries. It is but a small sample of what exists out there.
Extensive edits and additions were carried out by Rewan Demontay May 6, 2020.
Some follow-up mods by Laska subsequently: let's keep building it!